AFRL-SN-RS-TR-2007-145 Final Technical Report May 2007 WIDEBAND ELECTROABSORPTION MODULATOR FOR ANALOG APPLICATIONS University of California at San Diego

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.

STINFO COPY

AIR FORCE RESEARCH LABORATORY SENSORS DIRECTORATE ROME RESEARCH SITE

ROME, NEW YORK

NOTICE AND SIGNATURE PAGE Using Government drawings, specifications, or other data included in this document for any purpose other than Government procurement does not in any way obligate the U.S. Government. The fact that the Government formulated or supplied the drawings, specifications, or other data does not license the holder or any other person or corporation; or convey any rights or permission to manufacture, use, or sell any patented invention that may relate to them. This report was cleared for public release by the Air Force Research Laboratory Rome Research Site Public Affairs Office and is available to the general public, including foreign nationals. Copies may be obtained from the Defense Technical Information Center (DTIC) (http://www.dtic.mil). AFRL-SN-RS-TR-2007-145 HAS BEEN REVIEWED AND IS APPROVED FOR PUBLICATION IN ACCORDANCE WITH ASSIGNED DISTRIBUTION STATEMENT. FOR THE DIRECTOR: /s/ /s/ JAMES R. HUNTER RICHARD G. SHAUGNESSY Work Unit Manager Chief, Rome Operations Site Sensors Directorate This report is published in the interest of scientific and technical information exchange, and its publication does not constitute the Government’s approval or disapproval of its ideas or findings.

REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Washington Headquarters Service, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188) Washington, DC 20503. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. 1. REPORT DATE (DD-MM-YYYY)

MAY 2007 2. REPORT TYPE

Final 3. DATES COVERED (From - To)

Feb 06 – Feb 07 5a. CONTRACT NUMBER

5b. GRANT NUMBER FA8750-06-1-0055

4. TITLE AND SUBTITLE WIDEBAND ELECTROABSORPTION MODULATOR FOR ANALOG APPLICATIONS

5c. PROGRAM ELEMENT NUMBER 62500F

5d. PROJECT NUMBER WEMD

5e. TASK NUMBER SN

6. AUTHOR(S) P.K.L. Yu, I. Shubin, X.B. Xie and W.S.C. Chang

5f. WORK UNIT NUMBER 01

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) University of California at San Diego9500 Gilman Dr. Dept 621 La Jolla CA 92093-5004

8. PERFORMING ORGANIZATION REPORT NUMBER

10. SPONSOR/MONITOR'S ACRONYM(S)

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) AFRL/SNDP 25 Electronic Pky Rome NY 13441-4515

11. SPONSORING/MONITORING AGENCY REPORT NUMBER AFRL-SN-RS-TR-2007-145

12. DISTRIBUTION AVAILABILITY STATEMENT APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED. PA# 07-253

13. SUPPLEMENTARY NOTES

14. ABSTRACT There were two main technical objectives of this program with respect to the investigation of the high speed waveguide electroabsorption (EA) modulator, namely: (1) Design and fabrication of a waveguide modulator with a widened optical waveguide for easy packaging and lower insertion loss, and (2) interfacing with Infotonics for their fiber packaging effort of the modulator. In addition, an examination of the limits to Radio Frequency (RF) link gain, noise figure and spurious free dynamic range (SFDR) of the EA modulator was accomplished. This program produced the following accomplishments: 1) Finished a design for the modulator with large optical waveguide to improve the coupling in materials structure of either bulk InGaAsP or multiple quantum wells. The design has been fabricated at UCSD and repeated at a commercial foundry. 2) Examined the limits of RF link gain, noise figure and SFDR of EA modulators.

15. SUBJECT TERMS Optical modulator, Analog Optical Links, electroabsorption modulator, RF optical links

16. SECURITY CLASSIFICATION OF: 19a. NAME OF RESPONSIBLE PERSON James R. Hunter

a. REPORT U

b. ABSTRACT U

c. THIS PAGE U

17. LIMITATION OF ABSTRACT

UL

18. NUMBER OF PAGES

23 19b. TELEPHONE NUMBER (Include area code)

Standard Form 298 (Rev. 8-98)

Prescribed by ANSI Std. Z39.18

TABLE OF CONTENTS

1. Technical Summary ..................................................................................................... 1

2. Introduction.................................................................................................................. 1

3. Summary of accomplishments .................................................................................... 2

3.1 High power diluted waveguide electroabsorption modulator................................... 2

3.2 RF link gain and noise figure limits of electroabsorption modulator...................... 2

3.3 Limit of Spurious free dynamic range of electroabsorption modulator ................... 2

4. Technical progress achieved on project..................................................................... 3

4.1 High power diluted waveguide electroabsorption modulator................................... 3

4.2 RF link gain and noise figure limits of electroabsorption modulator...................... 7

4.3 Limit of Spurious free dynamic range of electroabsorption modulator ................. 11

5. Conclusion and future plan....................................................................................... 15

6. References. ................................................................................................................... 16

7. Glossary for Acronym ................................................................................................ 17

Appendix: Publications .................................................................................................. 18

i

LIST OF FIGURES

Figure 1 Device cross-section of the diluted waveguide EAM ......................................... 4

Figure 2 Schematic view of the EAM with the ground-signal electrode configuration. ... 5

Figure 3 Photograph of the wire-bonded EAM. ............................................................... 5

Figure 4 A practical equivalent circuit model of the EAM ............................................... 6

Figure 5 Calculated and measured μwave impedance for 2μm wide waveguide EAM..... 6

Figure 6 Small-signal ac equivalent circuit of electroabsorption modulator.. ................... 8

Figure 7 Link electrical gain as a function of laser power................................................. 9

Figure 8 Effect of equivalent Vπ on gain and noise figure............................................... 10

Figure 9 Experimental measurement of a link using an EAM modulator at 1550 nm .... 11

Figure 10 Calculated link output noise floor and multi-octave IIP3 of EAM. ................ 14

Figure 11 Calculated RF link gain, multi-octave link OIP3 and SFDR dependence of EAM on input optical power. .............................................................................................................. 15

LIST OF TABLES

Table 1. Material Layer structure of the diluted core waveguide electroabsorption

modulator………………………………………………………………………………….3

ii

1. Technical Summary:

There were two main technical objectives of this program with respect to the investigation of the

high speed waveguide electroabsorption (EA) modulator, namely: (1) Design and fabrication of a

waveguide modulator with widened optical waveguide for easy packaging and lower insertion

loss, and (2) interfacing with Infotonics for their fiber packaging effort of the modulator.

In addition, an examination of the limits to Radio Frequency (RF) link gain, noise figure and

spurious free dynamic range (SFDR) of the EA modulator was accomplished.

This report1 details the University of California at San Diego (UCSD) efforts in a multi-year

collaborative research program with AFRL at Rome Research Site who evaluated the EA

modulators in fiber links.

2. Introduction:

This program produced the following accomplishments:

1. Finished a design for the modulator with large optical waveguide to improve the coupling

in materials structure of either bulk InGaAsP or multiple quantum wells. The design has

been fabricated at UCSD and repeated at a commercial foundry. While the UCSD

fabrication run has contact adhesion problem, the fabrication run at the commercial

foundry has resolved this problem. The optical and electrical performances of devices

fabricated at both locations are similar.

2. Examined the limits of RF link gain, noise figure and SFDR of EA modulators.

1 The tasks reported in this project are also partially funded by program supported by DARPA and Air Force via Lockheed Martin.

1

3. Summary of accomplishments

3.1 High power diluted waveguide electroabsorption modulator A diluted waveguide electroabsorption modulator using bulk InGaAsP

electroabsorption layer has been designed and fabricated. The same design was

transferred to a commercial foundry. The resulting devices achieved the same device

properties as those fabricated at UCSD, with powers reaching 100 mW. A realistic

and practical microwave equivalent circuit model was also developed which explains

the S-parameters that were measured experimentally.

3.2 RF link gain and noise figure limits of electroabsorption modulator In a collaborative effort with Photonic Systems Inc., the RF link limit and noise figure

limit of electroabsorption modulator operating under optimal conditions has been

theoretically and experimentally established. The photocurrent at the modulator

results in a feedback effect that limits the available RF power reaching the modulator.

This limits the maximum link gain and noise figure of links using the

electroabsorption modulator.

3.3 Limit of Spurious free dynamic range of electroabsorption modulator The spurious free dynamic range (SFDR) of electroabsorption modulator has been

examined. The same current feedback mechanism that limits the RF link gain is found

to be beneficial for the high SFDR operation.

2

4. Technical progress achieved on project.

4.1 High power diluted waveguide electroabsorption modulator

The basic device structure of the electroabsorption modulator investigated under this

program is a continuation of the structure investigated under an Air Force program

(Wideband Agile Modulator) at Lockheed Martin with a subcontract at UCSD. The

frequency bandwidth is set at 20 GHz. The material structure is based upon the

bandwidth and modulation efficiency requirements. Initially the substrate is n-type

Indium Phosphide (InP) to ease the RF package requirements. It was later determined

that there is a considerable problem with the metal bonding on the p-electrode on top of

BCB. The recommendation (to the commercial foundry) was to make them on semi-

insulating InP. Table 1 summarizes the original material layer structure on n-InP

substrate design.

Table 1. Material Layer structure of the diluted core waveguide electroabsorption modulator. (The quaternary InGaAsP Q1.46 is the modulation layer; the Q1.15 is the waveguide layer, with bandgap wavelength of 1.46 μm and 1.15 μm respectively)

Material Thickness Depth Doping Index (1550 nm)

InGaAs 40 nm 0.04 μm P 1e19 3.6

InP 1 μm 1.04 μm P 4e17 3.16

InP 0.15 μm 1.19 μm undoped 3.17

Q1.46 0.275 μm 1.465 μm undoped 3.466

InP 0.01 μm 1.475 μm undoped 3.17

Q1.15 1.2 μm 2.675 μm N 1e18 3.31

InP 0.5 μm 3.175 μm N 1e18 3.16

InP 350 μm +/- 25 μm 3~8e18 3.17 To facilitate the optimal coupling to single mode fiber, an experiment was accomplished

to test the fiber coupling with different lensed fiber tip spot sizes. It was found that, for

3

the 2.5 μm mesa width, an optimal coupling was achieved with lensed fiber with a spot

size of larger than 4 μm. Thus the optimal coupling can be obtained with the mesa width

made closer to the same spot size. This is why the 4 μm wide mesa was chosen, as

shown in figure 1, for the Lockheed Martin devices. Because of the frequency

requirement, the capacitance of the modulator was limited by the mesa width and length,

as the thickness of the Q1.46 is separately determined from desired modulation

efficiency. Device lengths of 180 – 200 μm were targeted in the fabrication run. In order

to fit the microwave electrodes in a small space, the ground signal electrode configuration

was used, as depicted in figure 2. Figure 3 shows a photograph of the wire-bonded EAM.

InGaAs 1 x 1019 400Åindex 3.592

p type InP 4 x 1017 1.0 μm

Index 3.166

intrinsic-InP 1500Å

Intrinsic-InGaAs - Q1.46≤ 4 x 1016 , 0.275 μm,

Index 3.498

n type InP 1 x 1018 0.5 μm

S.I. type Substrate 3 × 1018

n type InGaAsP-Q1.15 1 x 1018 1.2 μm Index 3.29

InP etch stop 100 Å

4 μm

Figure 1 Device cross-section of the diluted waveguide EAM made from materials shown in Table 1

In the course of the project, the thickness of the Q1.46 was modified to 0.375 μm, while

the Q1.15 layer was increased to 2 μm, in order to better couple to the lensed fiber. The

resulting EAM has a reverse-breakdown voltage in excess of 15 V and a Vπ of 3 V for a

~200 mm long waveguide. The device can withstand up to 100 mW of input optical

4

power with a lowest fiber-to-fiber insertion of 5.5 dB, and a single-octave spurious free

dynamic range in excess of 120 dB in a one hertz bandwidth.2

140 ±10?m

300 μm(single device die)

180um+ -7um

Ground

Signal (bias) terminal

Figure 2 Schematic view of the EAM with the ground-signal electrode configuration.

Figure 3 Photograph of the wire-bonded EAM.

Modeling of the microwave electrode – It was determined that the text book type

microwave equivalent circuit model of the EAM waveguide does not work well for the

current devices due to the ideal circuit elements assumed in the model. Instead, to

2 EAM results measured at Lockheed Martin.

5

correlate the experimental measured microwave loss and index with the theory, there is a

need to develop an analysis that is based on the physical structure and materials

properties of the EAM.

Rcenterconductor metal

Rp-layer parallel Rn-layer parallel

Rground conductor metals Lmetal to n-layer

Rcontact Gi-layer conductivity (assumed to be small due to reverse bias and no photocurrent)

• Ci-layer

Rn-layer and p-layer perpendicular

Cmushroom cap

Figure 4 A practical equivalent circuit model of the EAM

This model, as depicted in figure 4, assumes that all of the components are at least, in

some capacity, functions of frequency based on the effects of skin effect in each of the

conductive materials. The calculated impedances for a 2 μm wide microwave waveguide

from the model match very well the measured values (see figures 5a and 5b).

Calculated Microwave Impedance

{Real part} {Imag . part}

0 2 4 6 8 1 1 1 1 1-

-

-

-

-

0

1

2

3

4

5

X: Y:

Frequency

X: Y: -

X: Y:

X: Y: -

X: Y: -

X: Y:

Calculated Microwave Impedance

0 2 4 6 8 1 1 1 1 1-50 -40 -30 -20 -10

0 10 20 30 40 50

Zo (Ohm)

Calculated Microwave Impedance

{Real part} {Imag . part}

(a)

X: 5.043 Y: 23.71

X: 10.01 Y: 23.27

X: 15.07 Y: 23.2

X: 5.043 Y: -5.561

X: 10.01 Y: -3.576

X: 15.07 Y: -3.266

0 2 4 6 8 1 1 1 1 1-30

-20

-10

0

10

20

30

4

X: Y:

Frequency

X: Y: -

X: Y:

X: Y: -

Measured Microwave Impedance

{Real part}

{Imag . part}

0 2 4 6 8 1 1 1 1 1

0

40

Zm

Measured Microwave Impedance

{Real part}

{Imag . part}

(b)

X: 15.07 Y: -0.3303

X: 15.07 Y: 24.33

X: 5.043 Y: -1.802

X: 5.043 Y: 21.32

Figure 5 Calculated and measured μ wave impedance for 2μm wide waveguide EAM

6

It should be noted the measured 2 μm wide microwave waveguide structure yields about

3dB/mm excluding the minor portion due to the probe pad transitions. This results in a

small microwave loss for the short EAM waveguide used.

4.2 RF link gain and noise figure limits of electroabsorption modulator

In an external modulation analog link, it is customary to treat the modulator as an ideal

three-terminal device where the light is controlled by the voltage applied to the

modulator, but there is no effect of the light on the voltage. This assumption is

appropriate for modulators where the modulation is based on the linear electro-optic

effect. For direct modulation links, the light is produced by the current supplied to the

transmitter laser, so there is a direct relation between electrical power supplied to the

optical transmitter and the light output. This results in a limitation on the gain of direct

modulation links that does not exist for external modulation links [1].

Electroabsorption modulators are intermediate between these two extremes. They are

external modulators and they affect the light through voltage-controlled absorption.

However, the absorption produces photocurrent, which interacts with the electrical

circuit. At low optical power the electroabsorption modulator behaves like an ideal

external modulator, but at high optical power it exhibits a gain limit.

This effect of photocurrent on gain was noticed when electroabsorption modulators began

to be able to handle optical powers of several mW [2]. This led to the observation that

there was a limit on the modulation efficiency of the electroabsorption modulator as the

optical power increased [3]. The origin of this limit and how it limits the performance of

analog links using electroabsorption modulators was investigated. Experimental data

confirming the link gain limit at very high optical power levels was shown.

The basis for this analysis is the equivalent circuit shown in figure 6 where the

photocurrent effect is represented by a resistor because it is a voltage-dependent current.

The ac voltage on the modulator is vm. The analysis is simplified by setting CM = 0.

7

Looking only at the low-frequency effects of the photocurrent the gain limit in its

simplest form can be seen. When CM ≠ 0, the photocurrent has additional effects such as

increasing the 3-dB bandwidth [2], but it does not change the basic effect.

Figure 6 Small-signal ac equivalent circuit of electroabsorption modulator. The resistor RP represents the voltage-dependent modulator photocurrent source.

This model considers only two sources of loss: voltage-independent coupling losses (tI

and tO), and voltage-dependent absorption loss. The optical power in the modulator input

waveguide is pIN = pLtI, where pL is the input laser power. The optical power in the

modulator output waveguide is pOUT. This analysis applies to small signals with ac

voltage much less than Vπ. The equivalent circuit can be solved to give the modulator

voltage vm in terms of the source current is as

⎟⎟⎠

⎞⎜⎜⎝

⎛+

+++

=

SL

SLM

MILSL

SLsm

RRRRR

VtpRR

RRiv

π

πη2

1

1 (1)

where RL is the modulator termination resistance, RS is the source impedance, RM is the

resistance in series with the modulator junction, and ηM is the modulator responsivity at

the bias point. The link gain is the ratio of the output RF power to the input RF power.

The input RF power is defined as the power delivered by the source to a matched load,

which is the available power ⟨is2⟩RS/4. The link output is the power delivered to the

detector load resistance RD.

8

Under the assumption that there are no losses in the link except the modulator, the link

gain is given by eq. 2,:

( )

2

2

2

21

142

⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢

⎣

⎡

⎟⎟⎠

⎞⎜⎜⎝

⎛+

++⎥⎦

⎤⎢⎣

⎡

+⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛=

SL

SLM

MILSL

SLLD

DOIL

RRRRR

VtpRR

RRRRVttpg

π

π πηπη (2)

where ηD is the detector responsivity. The gain is the product of three terms: the link

gain for an external modulation link with impedance-matched input, the effect of an

impedance mismatch between the source and termination, and a last term with the

dependence on the input optical power. In the limit of small pL, this term approaches

unity and the link behaves as expected for an external modulation link.

In the limit of large pL, the third term becomes inversely proportional to pL. In this limit

the gain becomes independent of either pL or Vπ, and is given by

2

2

1

4

⎟⎟⎠

⎞⎜⎜⎝

⎛++

⎟⎟⎠

⎞⎜⎜⎝

⎛=

L

M

S

M

S

D

M

DOLimit

RR

RR

RR

tgη

η (3)

Figure 7 Link electrical gain as a function of laser power, for various values of the modulator responsivity ηM (A/W). The dc component of the modulator photocurrent is also plotted. The parameter values are: Vπ = 1 V, RS = RL = RD = 50 Ω, RM = 5 Ω, ηD = 0.8 A/W, tI = -2 dB, tO = -2 dB, and tB = 0.5. [4] B

9

The effect of this gain limit is shown in figure 7 [4]. The case of ηM = 0 is the standard

external modulation result with no photocurrent effect. The case with ηM = 1 A/W

approximates performance expected from a high-power electroabsorption modulator. For

a high-performance modulator the limiting value is near 0 dB. The limit can be increased

if the modulator responsivity is reduced.

Figure 8 Effect of equivalent V on gain and noise figure. π The modulator responsivity is 1 A/W.

Other parameters are as in Fig. 7.

The gain limit also results in a minimum noise figure. The link electrical noise figure is

given by:

( ) ( )

o

LS

SLMSBMIL

o

DDOBILR

o

out

kTRR

RRRRttep

gkTRtttep

gf

gkTNf

2

1121

2

⎥⎦

⎤⎢⎣

⎡ ++−

+++==η

η (4)

where Nout is the total output noise, fR is the receiver noise figure (fR = 1 in this case), k is

Boltzmann’s constant, To is 290K, and e is the elementary charge. The first three terms

are the familiar input, receiver, and detector shot noise terms. The fourth term is due to

shot noise from the dc component of the modulator photocurrent. For small ηM or for

low bias (small tB) the modulator shot noise term becomes the dominant term at high

optical power. The noise figure is plotted in figure. 8.

B

10

The gain limit has been verified by measuring the gain of a link using an

electroabsorption modulator at high optical power levels. The modulator equivalent Vπ

was 0.85 V and the input and output losses were approximately tI = tO = 0.5. The bias

point was tB = 0.5, which occurred at 1.5 V reverse bias. The ac input voltage was 0.063

V peak-to-peak. The modulator’s apparent dc responsivity varied from 0.7 to 1.5 A/W,

indicating some mechanism creating additional photocurrent beyond simple absorption.

An RF responsivity, ηM = 0.8 A/W, was used to fit the calculation to the measured data.

The measurement frequency was 50 MHz, well below the RC bandwidth.

The results are shown in figure 9. The gain follows the theoretical prediction very

closely. The gain deviates from the prediction of this model only at the highest powers

used (>250 mW) due to heating.

Figure 9 Experimental measurement of a link using an EAM modulator at 1550 nm, compared with the theoretical gain calculation.

4.3 Limit of Spurious free dynamic range of electroabsorption modulator

The main conclusion from Section B.2 was that the voltage reduction across the p-i-n

junction, due to the negative feedback effect generated by the photocurrent, causes link

gain to deviate from the quadratic dependence on input optical power and finally

11

approach a gain limit. Here it is shown that the same mechanism affects EAM linearity

as well. The linearity performance solely due to the photocurrent feedback effect was

analyzed. The results show an input third-order intercept point (IIP3) dependence on

fourth-order power of optical power at substantially high power, which surpasses the

increase in noise which is linearly dependent on optical power. The SFDR of the link is

thus anticipated to improve with increased optical power.

While the input optical power is increased, both dc and ac photocurrent generation

increases as well. As a result, voltage drop on the source resistance RS and serial

resistance RM takes more portion of the total source voltage, leaving less modulation

voltage on the p-i-n junction. From a feedback point of view, what happens in the EAM

resembles a negative feedback system. The incoming voltage vS modulates the junction

and produces intensity modulation of optical carrier PLtItPtO[T(VB)-T(VB BB+vM)], where T(V)

is the optical transfer function of the EAM. At the same time, the modulated light

generates an ac photocurrent. The ac photocurrent effectively reduces the voltage across

the junction. This is effectively a negative feedback system with the output coupled into

the photodetector and generates an output voltage vL across load resistance RD. When the

input optical power is low, the photocurrent feedback can be ignored and the linearity of

electro-to-optical conversion is determined by the optical transfer function T(V). When

the optical power increases, the effective voltage across EAM junction is no longer vS,

but vM which is modified by photocurrent feedback.

By using a voltage gain function vOUT = g(vIN) without feedback, vIN and vOUT under

feedback can related as follows.

( ) OUTOUTIN vfvvg =− (5)

where g is a function including nonlinear harmonics caused by optical transfer curve

T(V), and f is the negative feedback coefficient related to the modulator parameter and the

remote resistor. It should be noted that vIN = vS/2 so as to conform with the conventional

definition of fiber-optic link gain, where the input RF power is taken with a modulator

12

load matched to the source [5].

The output voltage vOUT is equivalent to vL, the ac voltage across the load resistance of

the photodetector. It is well known in electronic amplifier design that negative feedback

can improve the linearity of the whole system if the feedback coefficient is more linear

than the transfer function of the system without feedback [6]. When the loop voltage

gain is large, the overall feedback system response is close to an inverted feedback

network response. In this case, function g includes optical transfer curve nonlinearities.

However, feedback coefficient f has nothing to do with the nonlinear transfer curve.

When the voltage gain without feedback is high enough, the voltage gain can be

approximated as 1/f. The system linearity is therefore determined mainly by f, not the

EAM transfer function. The EAM and photodetector responsivities ηM and ηD involved

in f can still affect the system linearity.

The impact of this fiber-optic link linearity by photocurrent feedback can also be

analyzed by separating intrinsic and extrinsic optical transfer curves. The intrinsic optical

transfer curve is defined as a function of junction voltage T(VM). It is clear that the

aforementioned optical transfer curve is equivalent to the intrinsic optical transfer curve

definition. Also the extrinsic optical transfer curve is then defined dependent on vIN =

vS/2, Te(VIN). The extrinsic optical transfer curve includes negative photocurrent

feedback effect and governs the linearity when the EAM gain is saturated. Different

orders of derivatives of both intrinsic and extrinsic optical transfer curves with respect to

their arguments can be evaluated and related based on the EAM equivalent circuit model;

for instance:

( )

MMSMPIL

M

IN

e

dVdTRRttP

dVdT

dVdT

+−=

η1

2 (6)

Equation (6) is the relationship between first order derivatives of the transfer curves. It

accounts for the link gain saturation. The term dT/dVM is considered negative due to the

fact that larger voltage causes less optical transmission. The denominator on the right

hand of (6) becomes much larger than unit when the input optical power is high enough,

which reduces the link gain. It can be lump as an EAM saturation factor k.

13

( ) ( )π

πηη

VRRtP

dVdTRRtPk MSMIL

MMSMIL 2

11+

+=+−= (7)

The derivatives of extrinsic and intrinsic optical transfer curves are related by a factor of

k3 for the second order, and k4 for the third order when the EAM is biased at its largest

slope efficiency voltage point where second order derivative nulls out. The second order

null point is also the bias point for multi-octave operation. It is clear that the derivatives

of extrinsic optical transfer curves become much smaller than that of intrinsic optical

transfer curves when saturation factor k>>1.

-20 -10 0 10 20 30

-180

-170

-160

-150

-140

-130

-120

Input Optical Power (dBm)

Noi

se F

loor

with

out R

IN (d

Bm

)

20

30

40

50

60

70

80

Multi-octave IIP

3 (dBm)

Figure 10 Calculated link output noise floor and multi-octave IIP3 of EAM as a function of input optical power. Laser RIN noise is not included. Low power EAM IIP3 of 20 dBm is assumed. Extra optical loss caused by dc bias of EAM is 3 dB. Other parameters used in the calculation are: tI = tO = 0.5, tP = -1 dB, RS = RD = 50 Ω, RM = 5 Ω, ηM = ηD = 1 A/W, Vπ = 1.5 V.

14

-20 -10 0 10 20 30-80

-70

-60

-50

-40

-30

-20

-10

0

-60

-40

-20

0

20

40

60

80

70

80

90

100

110

120

130

140

150

Input Optical Power (dBm)

RF Link Gain (dB) Multi-octave Link OIP3 (dBm) Multi-octave Link SFDR (dB/Hz2/3)

Figure 11 Calculated RF link gain, multi-octave link OIP3 and SFDR dependence of EAM on input optical power.

Thus it can be see that the IIP2 and IIP3 of a highly saturated EAM link can be improved

by a factor of k4 compared with non-feedback system, as illustrated in figures 10 and 11.

The output second and third-order intercept points (OIP2 and OIP3) increase by the same

factor as the gain saturates. On the other hand, the link output noise only increases

linearly with optical power, even when EAM shot noise dominates in the saturation case,

which is approximately proportional to k. Here laser relative intensity noise (RIN) is

excluded. Therefore link SFDR also improves by k2 under this situation.

5. Conclusion and future plan

Under the support of the Air Force Research Laboratory, much progress has been made

in this effort in the fabrication of the diluted core waveguide electroabsorption modulator

for analog fiber links. Notable progress has been made in the understanding of the

performance of electroabsorption modulator for analog applications. Further effort is

planned to show that RF fiber optic links using an EAM in the transmitter can achieve

large SFDR operation in a demonstration set-up.

15

6. References.

1. C.H. Cox III, Analog Optical Links: Theory and Practice, Cambridge University

Press, 2004, Ch. 3.

2. G.L. Li, C.K. Sun, S.A. Pappert, W.X. Chen, and P.K.L. Yu, “Ultrahigh-speed

traveling-wave electroabsorption modulator—design and analysis,” IEEE Trans.

Microwave Theory Tech., vol. 47, pp. 1177-1183, July 1999.

3. L.A. Johansson, Y.A. Akulova, G.A. Fish, and L.A. Coldren, “High optical power

electroabsorption waveguide modulator,” Electron. Lett., vol. 39, pp. 364-365, 20

Feb. 2003.

4. G. E. Betts, X. Xie, I. Shubin, W. S. C. Chang, and P. K. L. Yu, “Gain Limit in

Analog Links using Electroabsorption modulator,” IEEE Photonics Technology

Letters, Vol. 18, No. 19, pp. 2065-2067-1542, 2006.

5. C. H. Cox III, E. I. Ackerman, G. E. Betts, and J. L. Prince, “Limits on the

performance of RF-over-fiber links and their impact on device design,” IEEE

Trans. Microwave Theory and Tech., vol. 52, no. 4, pp. 906–920, Feb. 2006.

6. T. H. Lee, The Design of CMOS Radio-Frequency Integrated Circuits.

Cambridge, U.K.: Cambridge Univ. Press, 1998, ch. 14.

16

7. Glossary for Acronym

EAM = Electroabsorption modulator

IIP3 = input third-order intercept point

OIP3 = output third-order intercept point

RF = Radio Frequency

SFDR = Spurious Free Dynamic Range

UCSD = University of California, San Diego

Vπ = half wave voltage; voltage to generate a π phase shift

17

Appendix: Publications

1. G. E. Betts, X. Xie, I. Shubin, W. S. C. Chang, and P. K. L. Yu, “Gain Limit in

Analog Links using Electroabsorption modulator,” IEEE Photonics Technology

Letters, Vol. 18, No. 19, pp. 2065-2067-1542, 2006.

2. P. K. L. Yu, I. Shubin, X.B. Xie, W. S. C. Chang, “Recent Advances in Photonic

Devices for RF/Wireless Communication Applications” invited presentation, to

be presented at the 12th Microcoll Conference at Budapest, Hungary, May 2007.

3. X. B. Xie, I. Shubin, W. S. C. Chang, and P. K. L. Yu, “Analysis of Linearity of

Highly Saturated Electroabsorption Modulator Link due to Photocurrent Feedback

Effect”, to be submitted to Optics Express.

18